Allergen-Induced Airway Hyperresponsiveness Mediated by
Cyclooxygenase Inhibition Is Not Dependent on
5-Lipoxygenase or IL-5, but Is IL-13 Dependent1
R. Stokes Peebles, Jr.,2* Koichi Hashimoto,3* James R. Sheller,* Martin L. Moore,*
Jason D. Morrow,* Shaoquan Ji,†Jack A. Elias,‡Kasia Goleniewska,* Jamye O’Neal,*
Daphne B. Mitchell,* Barney S. Graham,§and Weisong Zhou*
Cyclooxygenase (COX) inhibition during allergic sensitization and allergen airway challenge results in augmented allergic in-
flammation. We hypothesized that this increase in allergic inflammation was dependent on increased generation of leukotrienes
that results from COX inhibition, as leukotrienes are important proinflammatory mediators of allergic disease. To test this
hypothesis, we allergically sensitized and challenged mice deficient in 5-lipoxygenase (5-LO). We found that 5-LO knockout mice
that were treated with a COX inhibitor during allergic sensitization and challenge had significantly increased airway hyperre-
sponsiveness (AHR) (p < 0.01) and airway eosinophilia (p < 0.01) compared with 5-LO knockout mice that were treated with
vehicle. The proinflammatory cytokines have also been hypothesized to be critical regulators of airway inflammation and AHR.
We found that the increase in airway eosinophilia seen with COX inhibition is dependent on IL-5, whereas the increase in AHR
is not dependent on this cytokine. In contrast, the COX inhibition-mediated increase in AHR is dependent on IL-13, but airway
eosinophilia is not. These results elucidate the pathways by which COX inhibition exerts a critical effect of the pulmonary
allergen-induced inflammatory response and confirm that COX products are important regulators of allergic inflammation. The
Journal of Immunology, 2005, 175: 8253–8259.
allergic phenotype occurs when mice are treated with a nonselec-
tive COX inhibitor or with pharmacologic agents that specifically
inhibit COX-1 or COX-2 (1–4). This up-regulation of the allergen-
induced pulmonary inflammatory response by COX inhibition is in-
dependent of signaling through the IL-4R or the STAT6 transcription
factor (1). Additionally, COX inhibition results in augmented aller-
gen-induced airway hyperresponsiveness (AHR) (2–4).
Currently, it is not known what factors regulate the heightened
allergic phenotype that is driven by COX inhibition. Arachidonic
acid can be metabolized not only through the COX enzymes, but
also through the 5-lipoxygenase (5-LO) pathway, an important
source of proinflammatory mediators (5). Leukotriene (LT)A4can
be metabolized by either LTA4hydrolase into LTB4, which is a
uring allergic sensitization and allergen airway chal-
lenge, cyclooxygenase (COX)4inhibition results in aug-
mented allergic inflammation (1–4). This increase in the
potent neutrophil chemoattractant and a weaker eosinophil chemo-
tactic agent, or by LTC4synthase into the cysteinyl LT (5). The
cysteinyl LT have eosinophil chemotactic properties, are potent
mucus secretagogues, and induce AHR (6). We have previously
shown that COX inhibition in the murine model of allergic pul-
monary inflammation results in increased cysteinyl LT levels (3).
Recent studies suggest that signaling through the LTB4receptor 1
is important in allergen-induced AHR (7, 8). Therefore, we hy-
pothesized that the increase in LT resulting from COX inhibition
mediates the augmented allergic phenotype in our model.
Cytokines are also important regulators of allergic inflammation
and AHR. IL-5 is a critical factor in eosinophil development, dif-
ferentiation, mobilization, activation, and survival (9–13). Animal
models of allergic sensitization in which IL-5 has either been neu-
tralized by Ab, or in which the gene producing IL-5 has been
deleted, suggest that IL-5 is critical to the control of airway eo-
sinophilia (14–16). The role of IL-5 in mediating AHR is less
clear, as some investigators have found that IL-5 is necessary for
AHR (16–19), whereas others have found that AHR can occur
independently of IL-5 (20–22). Mouse studies examining the role
of IL-13 suggest that it is necessary and sufficient to induce the
allergic phenotype, including AHR (23–25). Experimental work in
the murine model reveals that IL-13 acts on the resident cells in the
airway to induce mucus secretion and AHR, rather than through
recruitment of eosinophils to the lung or through type I-mediated
hypersensitivity (25). Therefore, we also sought to determine
whether the allergen-induced airway inflammation and AHR that
occurs with COX inhibition was dependent on either IL-5 or IL-13.
Materials and Methods
Pathogen-free 8- to 10-wk-old female mice were used in all experiments.
The 5-LO knockout (KO) mice and their wild-type (WT) control mice were
on a 129 genetic background and were purchased from The Jackson
*Department of Medicine, Vanderbilt University School of Medicine, Nashville, TN
37232;†LINCO Research, St. Charles, MO 63304;‡Department of Medicine, Yale
University, New Haven, CT 06520; and§Vaccine Research Center, National Institutes
of Health, Bethesda, MD 20892
Received for publication June 23, 2005. Accepted for publication September
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported by the American Academy of Allergy, Asthma and Im-
munology Education and Research Trust Grants and Awards HL-069949, AI 054660,
GM 15431, DK48831, and CA77839.
2Address correspondence and reprint requests to Dr. R. Stokes Peebles, Jr., Center for
Lung Research, T-1217 Medical Center North, Vanderbilt University Medical Center,
Nashville, TN 37232-2650. E-mail address: firstname.lastname@example.org
3Current address: Department of Microbiology, Fukushima University, Fukushima,
4Abbreviations used in this paper: COX, cyclooxygenase; 5-LO, 5-lipoxygenase;
AHR, airway hyperresponsiveness; BAL, bronchoalveolar lavage; indo, indometha-
cin; KO, knockout; LT, leukotriene; WT, wild type; PC, provocative concentration.
The Journal of Immunology
Copyright © 2005 by The American Association of Immunologists, Inc.0022-1767/05/$02.00
Laboratory. The IL-5 KO mice and their WT control mice were on a
C57BL/6 genetic background and were purchased from The Jackson Lab-
oratory. IL-13 KO mice were generated as previously described and had
been bred back to a C57BL/6 genetic background (26). In caring for ani-
mals the investigators adhered to the revised 1996 Guide for the Care and
Use of Laboratory Animals prepared by the Committee on Care and Use of
Laboratory Animals of the Institute of Laboratory Animal Resources, Na-
tional Research Council.
Allergen sensitization and challenge protocol
Mice were allergically sensitized by an i.p. injection of 0.1 ml (10 ?g) of
OVA (chicken OVA, grade V; Sigma-Aldrich) complexed with 20 mg of
Al(OH)3diluted in a volume of 100 ?l of PBS on day 0. On days 14
through 17, the mice were challenged by placing them in an acrylic box and
exposing them to aerosols of 1% OVA diluted in sterile PBS using an
ultrasonic nebulizer (Ultraneb 99; DeVilbiss) for 40 min each day. Non-
sensitized mice were injected i.p. with 20 mg of Al(OH)3diluted in a
volume of 100 ?l of PBS on day 0.
COX inhibitor administration
Indomethacin (30 ?g/ml) was administered in the drinking water starting
on day ?2 (2 days before the i.p. injection of OVA complexed with
Al(OH)3). An indomethacin stock was made by dissolving 150 mg of in-
domethacin in 50 ml of ethanol. Three times per week throughout the
experimental protocol, 2 ml of the indomethacin stock solution was added
to 200 ml of water in the animal water bottles. The water of the vehicle-
treated mice was also changed three times per week and 2 ml of ethanol
was added to 200 ml of water in the water bottles of those mice. Mice that
were allergically sensitized and challenged and treated with indomethacin
are designated OVA-indo. Mice that were allergically sensitized and chal-
lenged and treated with vehicle are designated OVA. Mice that were not
sensitized and treated with indomethacin are designated mock-indo. Mice
that were not sensitized and treated with vehicle are designated mock. We
have shown that this method of indomethacin administration significantly
reduces PGE2levels in bronchoalveolar lavage (BAL) fluid in the OVA
sensitization and challenge model (4).
Mice were anesthetized with i.p. injections of pentobarbital sodium (85
mg/kg) and a tracheostomy tube was placed. The internal jugular vein was
cannulated and a microsyringe was attached to i.v. tubing for methacholine
administration. The mice were then placed in a whole body plethysmog-
raphy chamber and mechanically ventilated. Lung resistance was measured
following administration of i.v. acetyl-?-methacholine (Sigma-Aldrich) as
previously described (27). Methacholine dose-response curves were ob-
tained by calculating the mean ? SE for individual animals at each metha-
choline dose. The provocative concentration dose of methacholine that led
to a 200% increase in baseline lung resistance (PC200) is expressed as
mean ? SE for individual animals.
Following lung resistance measurements, the animals were given a lethal
injection of pentobarbital. BAL was then performed by instilling 600 ?l of
5% BSA diluted in normal saline through the tracheostomy tube and then
withdrawing the fluid with gentle suction via the syringe. The typical BAL
fluid return was 300–400 ?l. White blood cells were counted on a hemo-
cytometer. Cytologic examination was performed on cytospin preparations
(Thermo Shandon). Cytospin slides were fixed and stained using DiffQuik
(American Scientific Products). Differential counts were based on counts of
100 cells using standard morphologic criteria to classify the cells as eo-
sinophils, lymphocytes, or other mononuclear leukocytes (alveolar macro-
phages and monocytes).
Measurement of cytokine proteins
Lung tissues were ground as previously described (4), and the levels of
IL-5 and IL-13 in the lung tissue homogenates were measured in the 129
WT and 5-LO KO mice with a commercially available LINCOplex mouse
cytokine/chemokine kit (Linco Research) using fluorescent-labeled micro-
sphere beads and a Luminex reader. Cytokine levels were measured in the
C57BL/6 WT and IL-5 KO mice using ELISA kits following the manu-
facturer’s instructions (R&D Systems). The experiments using 5-LO KO
and 129 mice were performed more recently and the newer fluorescent-
labeled microsphere bead technology was used in these experiments.
Before sacrifice on day 18, sera were collected from sensitized mice, then
analyzed by ELISA to determine levels of total IgE as previously described
(1). OVA-specific IgE levels were also measured by ELISA. In this pro-
tocol, 96-well Immunolon II plates (Nunc) were coated with 100 ?l of
OVA dissolved in sterile PBS at a concentration of 20 ?g/ml. Plates were
washed with PBS/0.5% Tween 20 and blocked with PBS/1% BSA for 1 h.
The plates were then washed before adding 100 ?l of serum diluted 1/200
in PBS/0.1% BSA/0.05% Tween 20 in duplicate. Plates were incubated
overnight and washed, and 100 ?l of rat anti-mouse IgE clone LO-ME-2
(Serotec) diluted 1/2000 was added to each well. After 1 h incubation at
37°C, the plates were again washed, and HRP activity was determined with
a tetramethylbenzidine (Sigma-Aldrich) developing solution (1% tetra-
methylbenzidine in DMSO, 0.001 M sodium acetate, and 0.45% H2O2final
concentration). Substrate development was stopped with 2.5 M H2SO4and
OD measured at 450 nm (OD450). The value reported for the OVA-specific
IgE concentration for each mouse is the average of duplicate OD450values.
Results are expressed as mean ? SEM. Measurements of cytokines by
ELISA, cell counts and differentials, PC200, and IgE by ELISA were ana-
lyzed by ANOVA with Fisher’s least significant difference performed as a
post hoc analysis. Dose-response curves to methacholine were compared
by repeated measures by ANOVA with Fisher’s least significant difference
performed as a post hoc analysis. Differences were considered to be sig-
nificant if values for p ? 0.05.
COX inhibition increased allergen-induced cellularity of BAL in
5-LO KO mice
Because cysteinyl LT are important mediators in allergic inflam-
mation, we measured BAL fluid cell counts and differentials in the
OVA and OVA-indo groups of WT and 5-LO KO mice to deter-
mine whether the increase in LT was responsible for the increased
allergic inflammation in indomethacin-treated mice. The WT
OVA-indo mice had greater total cells (50.6 ? 7.9 vs 25.3 ? 3.1 ?
105; p ? 0.01), eosinophils (42.3 ? 6.9 vs 20.2 ? 2.3 ? 105; p ?
0.01), and lymphocytes (2.0 ? 0.6 vs 0.6 ? 0.1 ? 105; p ? 0.01)
in the BAL fluid than did the WT OVA group or any other group
(Fig. 1). There were more total cells and eosinophils in the BAL
fluid of the 5-LO KO OVA-indo groups and WT OVA group than
in BAL fluids of the remaining groups, including the 5-LO KO
OVA group (p ? 0.01). Therefore, the increase in allergic inflam-
matory cell influx into the airways caused by COX inhibition is not
fully dependent on the expression of 5-LO.
COX inhibition increased AHR in allergen-sensitized
and -challenged 5-LO KO mice
Methacholine challenge was performed on day 18, one day after
the last OVA aerosol treatment in both WT and 5-LO KO mice.
The AHR was significantly greater in the 5-LO KO OVA-indo
mice compared with any other 5-LO KO group (p ? 0.05),
whereas the AHR of the 5-LO KO OVA mice was no different
from the 5-LO KO mock and mock-indo groups (Fig. 2). The AHR
was significantly greater in the WT OVA-indo group compared
with any other group of WT or 5-LO KO mice (p ? 0.05), whereas
the AHR of the WT OVA mice was greater than the AHR of the
WT mock and mock-indo groups (p ? 0.05), and no different from
the 5-LO KO OVA-indo group. The PC200values, reported as
micrograms of methacholine per gram of mouse, were 552 ? 184
for WT mock, 1039 ? 650 for WT mock-indo, 170 ? 21 for WT
OVA, and 162 ? 3 for WT OVA-indo, and there were no differ-
ences between the groups. The PC200values were 2792 ? 461 for
5-LO KO mock, 1460 ? 534 for 5-LO KO mock-indo, 1801 ?
413 for 5-LO KO OVA, and 678 ? 281 for 5-LO KO OVA-indo,
with values in the 5-LO KO OVA-indo group being significantly
82545-LO-INDEPENDENT ALLERGIC INFLAMMATION BY COX INHIBITION
lower than those in 5-LO KO mock (p ? 0.01) and 5-LO KO
OVA (p ? 0.05) groups.
The augmented IL-5 and IL-13 production resulting from COX
inhibition is dependent on 5-LO
Previously, we found that the levels of IL-5 in ground lung super-
natants peaked on day 16 of our protocol; therefore, we measured
cytokines on this day. We found that IL-5 (767 ? 218 vs 200 ?
126 pg/ml; p ? 0.01) and IL-13 (449 ? 105 vs 167 ? 61 pg/ml;
p ? 0.01) concentrations were greater in the lung supernatants of
the WT OVA-indo group compared with the WT OVA group (Fig.
3). However, there was only a trend for an increase in IL-5 (276 ?
123 vs 63 ? 8 pg/ml; p ? 0.06) and IL-13 (225 ? 67 vs 100 ?
5 pg/ml; p ? 0.1) in the 5-LO KO OVA-indo group compared with
the 5-LO KO OVA group that was not statistically significant. This
suggests that the increase in Th2 cytokines that occurs with COX
inhibition during the development of allergic inflammation is de-
pendent on 5-LO.
Elevated IgE levels resulting from COX inhibition is dependent
Because the WT OVA-indo and 5-LO KO OVA-indo mice had
greater numbers of inflammatory cells in the BAL fluid compared
with the rest of the mice in their respective groups, we sought to
determine whether this might be a result of an effect of COX in-
hibition on serum IgE levels (Fig. 4). We found that the WT OVA-
indo mice had greater serum concentrations of IgE compared with
WT OVA mice (30.3 ? 2.7 vs 13.7 ? 1.9 ?g/ml, respectively; p ?
0.01). However, there was no difference in the serum IgE levels
between the 5-LO KO OVA-indo and 5-LO KO OVA mice as
there were undetectable serum IgE levels in all the 5-LO KO OVA
mice and undetectable levels in four of the eight 5-LO KO OVA-
indo mice. We also measured OVA-specific IgE levels in the se-
rum of the mice and the results are reported as the OD450as stated
in Materials and Methods. The WT OVA-indo mice had signifi-
cantly greater serum concentrations of OVA-specific IgE com-
pared with WT OVA mice (304 ? 26 vs 91 ? 15 OD450; n ? 6
in each group; p ? 0.0001). OVA-specific IgE levels were unde-
tectable in the 5-LO OVA-indo and 5-LO OVA mice. Therefore,
this suggests that the increase in serum IgE levels seen with COX
inhibition is dependent on the presence of 5-LO.
COX inhibition mediated increase in BAL eosinophils is
dependent on IL-5, but not IL-13
IL-5 is a critical cytokine in eosinophil growth, differentiation, and
survival; however, IL-13 has also been reported to be a factor in
eosinophil recruitment to the airways (23). We therefore wanted to
determine whether the increased airway eosinophilia seen with
COX inhibition during the development of allergic pulmonary in-
flammation was mediated by IL-5 or IL-13. In these experiments,
we used WT, IL-5 KO, and IL-13 KO mice, all on a C57BL/6
background. We found that WT OVA-indo mice had increased
The data shown are a combination of two experiments (n ? 14–18 for
OVA-indo and OVA; n ? 5–8 for mock-indo and mock). Data points
represent the mean ? SEM measuring lung resistance. Statistical signifi-
cance is indicated by ?, p ? 0.05 compared with all other groups; †, p ?
0.01 compared with all other groups except WT OVA-indo and WT OVA;
and ‡, p ? 0.01 compared with all other groups except WT OVA-indo and
5-LO KO OVA-indo.
Results from methacholine challenge performed on day 18.
from WT and 5-LO KO mice on day 16 (n ? 4–5 for each group). Data
points represent the mean ? SEM and are representative of three separate
experiments. ?, p ? 0.01 compared with all other groups.
Concentrations of IL-5 and IL-13 in the lung supernatants
and lymphocytes in the BAL fluid on day 18 from al-
lergically sensitized and challenged or nonsensitized
WT and 5-LO KO mice treated with either the nonse-
lective COX inhibitor indomethacin (indo) or vehicle in
the drinking water (n ? 9–10 for all OVA and OVA-
indo groups and n ? 6–8 for all mock and mock-indo
groups). Statistical significance is indicated by ?, p ?
0.01 compared with all other groups; †, p ? 0.01 com-
pared with all other groups except WT OVA-indo and WT
OVA; and ‡, p ? 0.01 compared with all other groups
except WT OVA-indo and 5-LO KO OVA-indo. The re-
sults are representative of three separate experiments.
Total cells, macrophages, eosinophils,
8255The Journal of Immunology
total cells (20.4 ? 4.5 vs 8.8 ? 2.5 ? 105; p ? 0.05) and eosin-
ophils (15.4 ? 3.4 vs 5.6 ? 2.0 ? 105; p ? 0.05) in BAL fluid
compared with WT OVA mice (Fig. 5). In addition, we also found
that a similar number of total cells and eosinophils in the BAL
fluid of IL-13 KO OVA-indo mice compared with WT OVA-indo
mice, but the IL-13 KO OVA-indo mice had significantly greater
total cells (25.0 ? 5.8 vs 3.9 ? 0.6 ? 105; p ? 0.05) and eosin-
ophils (16.8 ? 4.7 vs 1.5 ? 0.5 ? 105; p ? 0.05) in BAL fluid
compared with the IL-13 KO OVA group. There was no difference
in the total cells (3.4 ? 1.1 vs 1.7 ? 0.4 ? 105) or eosinophils
(0.4 ? 0.1 vs 0.4 ? 0.3 ? 105) in the IL-5 KO OVA-indo mice
compared with the IL-5 KO OVA mice, and both of these groups
had significantly fewer total cells and eosinophils compared with
either the WT OVA-indo or IL-13 KO OVA-indo mice. Thus the
increase in BAL eosinophils seen with COX inhibition during the
development of allergic inflammation is mediated by IL-5, and not
The increased AHR seen with COX inhibition is dependent on
IL-13, but not IL-5
To determine whether the mechanism that led to an increase in
AHR seen with COX inhibition was driven by a cytokine that has
been reported to be important in AHR, we again used WT, IL-5
KO, and IL-13 KO mice, all on a C57BL/6 background. We found
that there was no difference in AHR between WT OVA, WT
mock-indo, and WT mock mice; however, the WT OVA-indo
group had significantly greater AHR (p ? 0.05) compared with the
other three groups (Fig. 6a). We also found similar results in the
IL-5 KO mice. There was no difference in AHR between the IL-5
KO OVA, mock-indo IL-5 KO OVA, and mock IL-5 KO OVA
mice; however, the IL-5 KO OVA-indo group had significantly
greater AHR (p ? 0.05) compared with the other three groups
(Fig. 6b). In contrast, we found there were no differences in AHR
among the IL-13 KO OVA-indo mice, IL-13 KO OVA mice,
IL-13 KO mock-indo mice, and IL-13 KO mock mice (Fig. 6c).
These results show, that despite an increase in BAL cell influx in
the IL-13 KO OVA-indo mice, the allergen-induced increase in
AHR with COX inhibition was IL-13 dependent.
We also calculated the PC200of methacholine for all the WT,
IL-5 KO, and IL-13 KO groups. For the C57BL/6 WT mice, the
PC200values, reported as micrograms of methacholine per gram of
mouse, were 405 ? 6 for WT mock, 347 ? 28 for WT mock-indo,
244 ? 39 for WT OVA, and 179 ? 51 for WT OVA-indo. The
PC200values for the WT OVA-indo and WT OVA groups were
each lower than either the WT mock or WT mock-indo (p ? 0.05),
but were not different from each other. For the IL-5 KO mice, the
PC200values were 387 ? 24 for the IL-5 KO-mock, 287 ? 58 for
the IL-5 KO-indo, 281 ? 44 for IL-5 KO OVA, and 160 ? 37 for
IL-5 KO OVA-indo. The IL-5 KO OVA-indo group had a signif-
icantly lower PC200compared with both the IL-5 KO OVA and
IL-5 KO mock groups (p ? 0.05), but there were no differences
between the other groups. For the IL-13 KO groups, the PC200
value was greater than 411 ?g/kg for all mice, except for two of
eight in the IL-13 KO OVA-indo group. There was no statistical
difference in PC200between the IL-13 KO groups.
COX inhibition increases IL-13 in WT and IL-5 KO mice, and
IL-5 in IL-13 KO mice, all on a C57BL/6 background
As AHR was increased in the WT OVA-indo and IL-5 KO OVA-
indo mice compared with these strains’ respective OVA groups,
yet AHR was not increased in the IL-13 KO OVA-indo group, we
sought to determine whether IL-13 levels in the lung were in-
creased in the WT OVA-indo and IL-5 KO OVA-indo mice. Cy-
tokine proteins were measured in the ground lung supernatants on
day 16 (Fig. 7). IL-5 was not detectable in the lung supernatants of
IL-5 KO mice, nor was IL-13 detectable in the lung supernatants
of the IL-13 KO mice. However, in WT mice, there was a trend
(p ? 0.1) for an increase in IL-5 in the OVA-indo mice (36.4 ?
15 pg/ml) compared with the OVA mice (?15.6 pg/ml, limit of
detection). IL-13 levels in the lung supernatants of the IL-5 KO
OVA-indo group were 174 ? 46 pg/ml compared with undetect-
able (?31.2 pg/ml) in the IL-5 KO OVA group (p ? 0.05). In WT
mice, the OVA-indo group had significantly greater levels of IL-13
in the lung supernatants compared with the WT OVA group
(110 ? 17 vs 28.4 pg/ml; p ? 0.05). Levels of IL-13 were not
detected in the lung supernatants of either nonsensitized IL-5 KO
mice or nonsensitized WT mice, whether treated with indometh-
acin or not. There was significantly greater IL-5 levels in the lungs
of the IL-13 KO OVA-indo group compared with the IL-13 KO
OVA mice (171 ? 28 vs ?15.6 pg/ml, limit of detection; p ?
ophils (eos), and lymphocytes (lymphs) in the BAL fluid
on day 18 from allergically sensitized and nonsensitized
WT, IL-5 KO, and IL-13 KO mice either treated with
the nonselective COX inhibitor indomethacin (indo) or
vehicle in the drinking water. Statistical significance is
indicated by ?, p ? 0.05 compared with the respective
OVA, mock, and mock-indo groups for that particular
strain and †, p ? 0.05 compared with the respective
mock and mock-indo groups for that particular strain.
Total cells, macrophages (macs), eosin-
mice at day 18 (n ? 8 for each group). Data are representative of three
separate experiments. ?, p ? 0.01 compared with all other groups.
Total IgE levels in serum harvested from WT and 5-LO KO
8256 5-LO-INDEPENDENT ALLERGIC INFLAMMATION BY COX INHIBITION
In our previous studies, we found that COX inhibition up-regulated
the allergic phenotype in a mouse model of allergen-induced in-
flammation (1, 3, 4). Others have found that deficiency of COX-1,
but not COX-2 deficiency, results in increased Th2 cytokine pro-
duction, AHR, and LT generation (28, 29). COX inhibition en-
hances the production of LT, presumably by decreasing PGE2, a
mediator that negatively regulates LT synthesis (30, 31). An al-
ternative theory is that COX inhibition prevents prostanoid forma-
tion and that the increased arachidonic acid substrate is therefore
“shunted” toward LT synthesis. However, there is experimental
evidence suggesting that shunting does not occur (32). Murine
models reveal that LT are important regulators of pulmonary al-
lergic inflammation. Mice lacking a functioning 5-LO enzyme had
significantly decreased airway eosinophilia, serum IgE levels, and
AHR compared with WT mice (33). Specific inhibitors of 5-LO
and 5-LO-activating protein decreased airway and lung eosino-
philia and airway mucus expression (34). Therefore, an important
question to be answered was whether the increase in allergic in-
flammation that occurs with COX inhibition seen in our model is
a result of an increase in LT generation. We found that the COX
inhibition-induced increase in allergic inflammation and AHR was
not solely dependent upon LT expression, as the 5-LO KO OVA-
indo mice had significantly increased airway inflammation and
AHR compared with the increase for the 5-LO KO OVA group.
We reported AHR by two methods: 1) dose-response curves of
lung resistance to increasing doses of methacholine and 2) the
PC200. The changes in AHR as measured by the methacholine
dose-response curve and the PC200were not entirely consistent.
This is a reflection of the fact that the differences between groups
observed in the methacholine dose-response curves occurred at
higher doses of methacholine. Both the methacholine dose-re-
sponse curves and PC200values are measures of AHR and one is
not necessarily more valid than the other. In this report, we use the
The increases in airway eosinophilia and AHR in the 5-LO KO
OVA-indo mice were not as great as was seen in the WT OVA-
indo mice, thus not ruling out the possibility that LT may be play-
ing some role in the COX inhibition-mediated enhancement of
allergic inflammation. Lung IL-5 and IL-13 expression was sig-
nificantly increased in the WT OVA-indo mice compared with the
WT OVA group, and although lung IL-5 and IL-13 expression was
3- and 2-fold greater in the 5-LO KO OVA-indo mice compared
with the 5-LO KO OVA group, these differences were not statis-
tically significant, suggesting that regulation of these cytokines
may be partially dependent on 5-LO expression. There was no
difference in total serum IgE levels in the 5-LO KO OVA-indo and
5-LO KO OVA mice, whereas there was a significant increase in
the WT OVA-indo group compared with the WT OVA mice, sug-
gesting that the COX inhibition-mediated increase in serum IgE is
dependent on the expression of 5-LO.
Because the augmentation of allergic inflammation by COX in-
hibition is not solely dependent on increased LT synthesis, this
suggests that a COX product restrains allergic inflammation inde-
pendently of an effect on LT expression. Murine models suggest
that several prostanoids may regulate allergic inflammation. Mice
lacking the PGE2receptor EP3 had increased allergic inflamma-
tion compared with either WT mice or mice lacking each of the
other three EP receptors (35), suggesting that signaling through
a, In the WT mice (n ? 6–9 mice per group) data are representative of two
separate experiments. b, Data shown for IL-5 KO are combined data from
two experiments (n ? 11 for OVA, n ? 12 for OVA-indo, n ? 5 from
mock, n ? 6 for mock-indo). c, In IL-13 KO mice (n ? 7–9 for the OVA
and OVA-indo groups and 4–6 for the mock and mock-indo groups), data
are representative of two separate experiments. Data points represent the
mean ? SEM measuring lung resistance. ?, p ? 0.05 compared with the
OVA, mock-indo, and mock groups.
Results from methacholine challenge performed on day 18.
from WT and IL-5 KO mice on day 16 (n ? 4–5 for each group). Data
points represent the mean ? SEM and are representative of three separate
experiments. ?, p ? 0.01 compared with all other groups.
Concentrations of IL-5 and IL-13 in the lung supernatants
8257 The Journal of Immunology
EP3 down-regulates allergic inflammation. Mice that are deficient
in the PGI2receptor IP had enhanced allergic inflammation, both
in acute and chronic models of allergen-induced pulmonary dis-
ease (36, 37), suggesting that PGI2also suppresses the allergen-
induced inflammatory response. In contrast, in in vivo models,
there is the suggestion that PGD2up-regulates the allergen-in-
duced inflammatory response. Mice lacking the PGD2receptor DP
had diminished pulmonary allergic inflammation (38), whereas
mice that overexpress PGD2synthase had augmented an aug-
mented allergic phenotype (39). Studies using either a thrombox-
ane synthase inhibitor or a thromboxane receptor antagonist de-
creased allergic inflammation (40), suggesting that this mediator
also enhances allergic inflammation in the lung. Given that the
overall effect of inhibiting COX increases allergic inflammation, it
is likely that the prostanoids that restrain allergic inflammation
have greater activity than those that increase the allergic
Cytokines are also very important mediators of allergic inflam-
mation. We previously found that the increase in the pulmonary
airway eosinophilia modulated by COX inhibition was indepen-
dent of signaling through either the IL-4R? or STAT6 (1). How-
ever, the pathway by which COX inhibition increases airway eo-
sinophilia is unknown. IL-5 is a critical factor in eosinophil
development, differentiation, mobilization, activation, and survival
(9–13). In vitro, IL-5 primes eosinophils to respond to eotaxin and
the ability of eotaxin to mediate eosinophil chemotaxis is aug-
mented by increased levels of IL-5 (41). We found that the in-
crease in airway eosinophilia was dependent on IL-5 expression. In
WT OVA-indo mice on a C57BL/6 background, there was a sig-
nificant increase in airway inflammation and eosinophilia com-
pared with WT OVA mice, which was absent in IL-5 KO OVA-
indo mice. We found that the COX inhibition driven increase in
airway eosinophils was not dependent on IL-13, as IL-13 KO
OVA-indo mice had significantly greater airway inflammation
compared with IL-13 KO OVA mice and was comparable to the
inflammation seen in WT OVA-indo mice. This finding would be
consistent with the report that IL-13 induced eosinophil recruit-
ment into thelungby an
Although we found that the increase in airway eosinophilia re-
sulting from COX inhibition was dependent on IL-5, we found that
the increase in allergen-induced AHR mediated by COX inhibition
was not dependent on this cytokine, but instead was dependent on
IL-13. There is considerable debate over the necessity for eosin-
ophils in the pathogenesis of airway reactivity (43–45). Although
some studies in mice have shown that IL-5 is essential for the
development of AHR (16–19), others have not (21, 22, 46). In our
experiments, we found that the WT OVA-indo group had signif-
icantly higher AHR compared with the WT OVA, WT mock-indo,
and WT mock groups. There was a small increase in the AHR of
the WT OVA group compared with the WT mock group, but this
increase was statistically insignificant. This diminished AHR re-
sponse in the WT OVA group compared with the nonsensitized
control may reflect a strain influence on the development of AHR,
as C57BL6 mice have diminished airway reactivity as compared
with BALB/c mice (47, 48). The IL-5 KO OVA-indo group had
significantly greater airway reactivity compared with the IL-5 KO
OVA group and the two IL-5 KO nonsensitized groups, and these
three groups were not statistically significantly different from each
other. These results show that COX inhibition during allergic sen-
sitization has profound effects on airway physiology in WT and
IL-5 KO mice that are on a C57BL/6 background. Our results
further show that IL-13 is responsible for the increased AHR in
both the WT and IL-5 KO OVA-indo groups, as lung expression
of IL-13 was significantly increased in these two groups compared
with their respective OVA groups and IL-13 KO OVA-indo mice
did not have significantly greater AHR compared with either IL-13
KO OVA mice or either of the two nonsensitized groups.
These findings are consistent with several studies examining the
role of IL-13 on airway inflammation and pulmonary physiology.
In a mouse model, anti-IL-13 Ab treatment significantly reduced
AHR while having no effect on eosinophilia (49). In this study,
combined administration of anti-IL-5 and anti-eotaxin Abs before
allergen challenge inhibited airway eosinophilia, but did not alter
AHR (49). Intratracheal administration of rIL-13 induced AHR
from 6 to 24 h after administration, and this increase in AHR was
dissociated from airway eosinophilia. These investigators also
found that the IL-13-mediated increase in AHR was independent
of IL-5 and eotaxin (50). Additionally, adoptive transfer of Th2
polarized cells from IL-13 KO mice that make high levels of IL-4
and IL-5 led to airway inflammation, yet failed to induce AHR
(51), and this induction of AHR was dependent on signaling
through IL-4R? (52). Taken together, these studies show that
IL-13 critically regulates AHR, without much effect on airway
In summary, we found that LT are not necessary for the increase
in the allergen-induced inflammatory response and AHR that re-
sults from COX inhibition during the development of allergic in-
flammation. We also found that the increase in airway eosinophilia
seen with COX inhibition is dependent on IL-5 whereas the in-
crease in AHR is not dependent on this cytokine. In contrast, the
COX inhibition-mediated increase in AHR is dependent on IL-13,
but airway eosinophilia is not. These results confirm that COX
products are important regulators of allergic inflammation.
The authors have no financial conflict of interest.
1. Hashimoto, K., J. R. Sheller, J. D. Morrow, R. D. Collins, K. Goleniewska, J.
O’Neal, W. Zhou, S. Ji, D. B. Mitchell, B. S. Graham, et al. 2005. Cyclooxy-
genase inhibition augments allergic inflammation through CD4-dependent,
STAT6-independent mechanisms. J. Immunol. 174: 525–532.
2. Jaffar, Z., K. S. Wan, and K. Roberts. 2002. A key role for prostaglandin I2in
limiting lung mucosal Th2, but not Th1, responses to inhaled allergen. J. Immu-
nol. 169: 5997–6004.
3. Peebles, R. S., Jr., R. Dworski, R. D. Collins, A. K. Jarzecka, D. B. Mitchell,
B. S. Graham, and J. R. Sheller. 2000. Cyclooxygenase inhibition increases in-
terleukin 5 and interleukin 13 production and airway hyperresponsiveness in
allergic mice. Am. J. Respir. Crit. Care Med. 162: 676–681.
4. Peebles, R. S., Jr., K. Hashimoto, J. D. Morrow, R. Dworski., R. D. Collins,
Y. Hashimoto, J. W. Christman, K. H. Kang, K. Jarzecka, J. Furlong, et al. 2002.
Selective cyclooxygenase-1 and -2 inhibitors each increase allergic inflammation
and airway hyperresponsiveness in mice. Am. J. Respir. Crit. Care Med. 165:
5. Smith, W. L., and D. L. Dewitt. 1996. Prostaglandin endoperoxide H synthases-1
and -2. Adv. Immunol. 62: 167–215.
6. Holgate, S. T., M. Peters-Golden, R. A. Panettieri, and W. R. Henderson, Jr.
2003. Roles of cysteinyl leukotrienes in airway inflammation, smooth muscle
function, and remodeling. J. Allergy Clin. Immunol. 111: S18–S34.
7. Miyahara, N., K. Takeda, S. Miyahara, S. Matsubara, T. Koya, A. Joetham,
E. Krishnan, A. Dakhama, B. Haribabu, and E. W. Gelfand. 2005. Requirement
for leukotriene B4 receptor 1 in allergen-induced airway hyperresponsiveness.
Am. J. Respir. Crit. Care Med. 172: 161–167.
8. Miyahara, N., K. Takeda, S. Miyahara, C. Taube, A. Joetham, T. Koya,
S. Matsubara, A. Dakhama, A. M. Tager, A. D. Luster, et al. 2005. Leukotriene
B4 receptor-1 is essential for allergen-mediated recruitment of CD8?T cells and
airway hyperresponsiveness. J. Immunol. 174: 4979–4984.
9. Clutterbuck, E. J., E. M. Hirst, and C. J. Sanderson. 1989. Human interleukin-5
(IL-5) regulates the production of eosinophils in human bone marrow cultures:
comparison and interaction with IL-1, IL-3, IL-6, and GMCSF. Blood 73:
10. Lopez, A. F., C. J. Sanderson, J. R. Gamble, H. D. Campbell, I. G. Young, and
M. A. Vadas. 1988. Recombinant human interleukin 5 is a selective activator of
human eosinophil function. J. Exp. Med. 167: 219–224.
11. Sanderson, C. J. 1988. Interleukin-5: an eosinophil growth and activation factor.
Dev. Biol. Stand. 69: 23–29.
12. Sanderson, C. J. 1990. The biological role of interleukin 5. Int. J. Cell Cloning
8(Suppl. 1): 147–153.
82585-LO-INDEPENDENT ALLERGIC INFLAMMATION BY COX INHIBITION
13. Wang, J. M., A. Rambaldi, A. Biondi, Z. G. Chen, C. J. Sanderson, and
A. Mantovani. 1989. Recombinant human interleukin 5 is a selective eosinophil
chemoattractant. Eur. J. Immunol. 19: 701–705.
14. Dent, L. A., M. Strath, A. L. Mellor, and C. J. Sanderson. 1990. Eosinophilia in
transgenic mice expressing interleukin 5. J. Exp. Med. 172: 1425–1431.
15. Foster, P. S., S. P. Hogan, A. J. Ramsay, K. I. Matthaei, and I. G. Young. 1996.
Interleukin 5 deficiency abolishes eosinophilia, airways hyperreactivity, and lung
damage in a mouse asthma model. J. Exp. Med. 183: 195–201.
16. Lee, J. J., M. P. McGarry, S. C. Farmer, K. L. Denzler, K. A. Larson,
P. E. Carrigan, I. E. Brenneise, M. A. Horton, A. Haczku, E. W. Gelfand, et al.
1997. Interleukin-5 expression in the lung epithelium of transgenic mice leads to
pulmonary changes pathognomonic of asthma. J. Exp. Med. 185: 2143–2156.
17. Hogan, S. P., A. Koskinen, and P. S. Foster. 1997. Interleukin-5 and eosinophils
induce airway damage and bronchial hyperreactivity during allergic airway in-
flammation in BALB/c mice. Immunol. Cell Biol. 75: 284–288.
18. Karras, J. G., K. McGraw, R. A. McKay, S. R. Cooper, D. Lerner, T. Lu,
C. Walker, N. M. Dean, and B. P. Monia. 2000. Inhibition of antigen-induced
eosinophilia and late phase airway hyperresponsiveness by an IL-5 antisense
oligonucleotide in mouse models of asthma. J. Immunol. 164: 5409–5415.
19. Shardonofsky, F. R., J. Venzor, III, R. Barrios, K. P. Leong, and D. P. Huston.
1999. Therapeutic efficacy of an anti-IL-5 monoclonal antibody delivered into the
respiratory tract in a murine model of asthma. J. Allergy Clin. Immunol. 104:
20. Corry, D. B., G. Gru ¨nig, H. Hadeiba, V. P. Kurup, M. L. Warnock, D. Sheppard,
D. M. Rennick, and R. M. Locksley. 1998. Requirements for allergen-induced
airway hyperreactivity in T and B cell-deficient mice. Mol. Med. 4: 344–355.
21. Hogan, S. P., K. I. Matthaei, J. M. Young, A. Koskinen, I. G. Young, and
P. S. Foster. 1998. A novel T cell-regulated mechanism modulating allergen-
induced airways hyperreactivity in BALB/c mice independently of IL-4 and IL-5.
J. Immunol. 161: 1501–1509.
22. Mathur, M., K. Herrmann, X. Li, Y. Qin, J. Weinstock, D. Elliott, J. Monahan,
and P. Padrid. 1999. TRFK-5 reverses established airway eosinophilia but not
established hyperresponsiveness in a murine model of chronic asthma.
Am. J. Respir. Crit. Care Med. 159: 580–587.
23. Gru ¨nig, G., M. Warnock, A. E. Wakil, R. Venkayya, F. Brombacher,
D. M Rennick, D. Sheppard, M. Mohrs, D. D. Donaldson, R. M. Locksley, et al.
1998. Requirement for IL-13 independently of IL-4 in experimental asthma. Sci-
ence 282: 2261–2263.
24. Wills-Karp, M., J. Luyimbazi, X. Xu, B. Schofield, T. Y. Neben, C. L. Karp, and
D. D. Donaldson. 1998. Interleukin-13: central mediator of allergic asthma. Sci-
ence 282: 2258–2261.
25. Wills-Karp, M. 2004. Interleukin-13 in asthma pathogenesis. Immunol. Rev. 202:
26. McKenzie, G. J., C. L. Emson, S. E. Bell, S. Anderson, P. Fallon, G. Zurawski,
R. Murray, R. Grencis, and A. N. McKenzie. 1998. Impaired development of Th2
cells in IL-13-deficient mice. Immunity 9: 423–432.
27. Chen, X.-S., J. R. Sheller, E. N. Johnson, and C. D. Funk. 1994. Role of leuko-
trienes revealed by targeted disruption of the 5-lipoxygenase gene. Nature 372:
28. Carey, M. A., D. R. Germolec, J. A. Bradbury, R. A. Gooch, M. P. Moorman,
G. P. Flake, R. Langenbach, and D. C. Zeldin. 2003. Accentuated T helper type
2 airway response after allergen challenge in cyclooxygenase-1?/?but not cy-
clooxygenase-2?/?mice. Am. J. Respir. Crit. Care Med. 167: 1509–1515.
29. Gavett, S. H., S. L. Madison, P. C. Chulada, P. E. Scarborough, W. Qu,
J. E. Boyle, H. F. Tiano, C. A. Lee, R. Langenbach, V. L. Roggli, et al. 1999.
Allergic lung responses are increased in prostaglandin H synthase-deficient mice.
J. Clin. Invest. 104: 721–732.
30. Docherty, J. C., and T. W. Wilson. 1987. Indomethacin increases the formation
of lipoxygenase products in calcium ionophore stimulated human neutrophils.
Biochem. Biophys. Res. Commun. 148: 534–538.
31. Elliott, G. R., A. P. Lauwen, and I. L. Bonta. 1989. Prostaglandin E2inhibits and
indomethacin and aspirin enhance, A23187-stimulated leukotriene B4 synthesis
by rat peritoneal macrophages. Br. J. Pharmacol. 96: 265–270.
32. Peters, S. P., D. W. MacGlashan, Jr., R. P. Schleimer, E. C. Hayes,
N. F. Adkinson, Jr., and L. M. Lichtenstein. 1985. The pharmacologic modulation
of the release of arachidonic acid metabolites from purified human lung mast
cells. Am. Rev. Respir. Dis. 132: 367–373.
33. Irvin, C. G., Y. P. Tu, J. R. Sheller, and C. D. Funk. 1997. 5-Lipoxygenase
products are necessary for ovalbumin-induced airway responsiveness in mice.
Am. J. Physiol. 272: L1053–L1058.
34. Henderson, W. R., Jr., D. B. Lewis, R. K. Albert, Y. Zhang, W. J. Lamm,
G. K. Chiang, F. Jones, P. Eriksen, Y. T. Tien, M. Jonas, et al. 1996. The im-
portance of leukotrienes in airway inflammation in a mouse model of asthma.
J. Exp. Med. 184: 1483–1494.
35. Kunikata, T., H. Yamane, E. Segi, T. Matsuoka, Y. Sugimoto, S. Tanaka,
H. Tanaka, H. Nagai, A. Ichikawa, and S. Narumiya. 2005. Suppression of al-
lergic inflammation by the prostaglandin E receptor subtype EP3. Nat. Immunol.
36. Nagao, K., H. Tanaka, M. Komai, T. Masuda, S. Narumiya, and H. Nagai. 2003.
Role of prostaglandin I2in airway remodeling induced by repeated allergen chal-
lenge in mice. Am. J. Respir. Cell Mol. Biol. 29: 314–320.
37. Takahashi, Y., S. Tokuoka, T. Masuda, Y. Hirano, M. Nagao, H. Tanaka,
N. Inagaki, S. Narumiya, and H. Nagai. 2002. Augmentation of allergic inflam-
mation in prostanoid IP receptor deficient mice. Br. J. Pharmacol. 137: 315–322.
38. Matsuoka, T., M. Hirata, H. Tanaka, Y. Takahashi, T. Murata, K. Kabashima,
Y. Sugimoto, T. Kobayashi, F. Ushikubi, Y. Aze. et al. 2000. Prostaglandin D2
as a mediator of allergic asthma. Science 287: 2013–2017.
39. Fujitani, Y., Y. Kanaoka, K. Aritake, N. Uodome, K. Okazaki-Hatake, and
Y. Urade. 2002. Pronounced eosinophilic lung inflammation and Th2 cytokine
release in human lipocalin-type prostaglandin D synthase transgenic mice. J. Im-
munol. 168: 443–449.
40. Shi, H., A. Yokoyama, N. Kohno, Y. Hirasawa, K. Kondo, K. Sakai, and
K. Hiwada. 1998. Effect of thromboxane A2 inhibitors on allergic pulmonary
inflammation in mice. Eur. Respir. J. 11: 624–629.
41. Lamkhioued, B., P. M. Renzi, S. Abi-Younes, E. A. Garcia-Zepada,
Z. Allakhverdi, O. Ghaffar, M. D. Rothenberg, A. D. Luster, and Q. Hamid. 1997.
Increased expression of eotaxin in bronchoalveolar lavage and airways of asth-
matics contributes to the chemotaxis of eosinophils to the site of inflammation.
J. Immunol. 159: 4593–4601.
42. Pope, S. M., E. B. Brandt, A. Mishra, S. P. Hogan, N. Zimmermann,
K. I. Matthaei, P. S. Foster, and M. E. Rothenberg. 2001. IL-13 induces eosin-
ophil recruitment into the lung by an IL-5- and eotaxin-dependent mechanism.
J. Allergy Clin. Immunol. 108: 594–601.
43. Hamelmann, E., and E. W. Gelfand. 1999. Role of IL-5 in the development of
allergen-induced airway hyperresponsiveness. Int. Arch. Allergy Immunol. 120:
44. Hamelmann, E., and E. W. Gelfand. 2001. IL-5-induced airway eosinophilia: the
key to asthma? Immunol. Rev. 179: 182–191.
45. O’Byrne, P. M., M. D. Inman, and K. Parameswaran. 2001. The trials and trib-
ulations of IL-5, eosinophils, and allergic asthma. J. Allergy Clin. Immunol. 108:
46. Corry, D. B., H. G. Folkesson, M. L. Warnock, D. J. Erle, M. A. Matthay,
J. P. Wiener-Kronish, and R. M. Locksley. 1996. Interleukin 4, but not interleukin
5 or eosinophils, is required in a murine model of acute airway hyperreactivity.
J. Exp. Med. 183: 109–117.
47. Brewer, J. P., A. B. Kisselgof, and T. R. Martin. 1999. Genetic variability in
pulmonary physiological, cellular, and antibody responses to antigen in mice.
Am. J. Respir. Crit. Care Med. 160: 1150.
48. Herz, U., A. Braun, R. Ruckert, and H. Renz. 1998. Various immunological
phenotypes are associated with increased airway responsiveness. Clin. Exp. Al-
lergy 28: 625.
49. Eum, S. Y., K. Maghni, B. Tolloczko, D. H. Eidelman, and J. G. Martin. 2005.
IL-13 may mediate allergen-induced hyperresponsiveness independently of IL-5
or eotaxin by effects on airway smooth muscle. Am. J. Physiol. 288: L576.
50. Yang, M., S. P. Hogan, P. J. Henry, K. I. Matthaei, A. N. McKenzie, I. G. Young,
M. E. Rothenberg, and P. S. Foster. 2001. Interleukin-13 mediates airways hy-
perreactivity through the IL-4 receptor-? chain and STAT-6 independently of
IL-5 and eotaxin. Am. J. Respir. Cell Mol. Biol. 25: 522.
51. Walter, D. M., J. J. McIntire, G. Berry, A. N. McKenzie, D. D. Donaldson,
R. H. DeKruyff, and D. T. Umetsu. 2001. Critical role for IL-13 in the develop-
ment of allergen-induced airway hyperreactivity. J. Immunol. 167: 4668.
52. Mattes, J., M. Yang, A. Siqueira, K. Clark, J. MacKenzie, A. N. McKenzie,
D. C. Webb, K. I. Matthaei, and P. S. Foster. 2001. IL-13 induces airways hy-
perreactivity independently of the IL-4R? chain in the allergic lung. J. Immunol.
8259The Journal of Immunology